A GIS-BASED APPROACH TO MODELING THREE-DIMENSIONAL GEOLOGY OF NEAR-SURFACE GLACIAL MORPHOSEQUENCES: HUNTERTOWN FORMATION, NORTHEASTERN INDIANA

Planners, managers, and groundwater modelers all need detailed information about the three-dimensional (3-D) distribution of sediments in the subsurface. Traditional methods of mapping (in two dimensions supplemented by idealized cross sections) fall short of these needs. As part of the Great Lakes Geologic Mapping Coalition, we are attempting to develop reproducible and verifiable 3-D models of sediment distribution in former glacial depositional settings.

The Huntertown Formation (Quaternary) in Allen County, Indiana, is located in an interlobate landscape composed of coarse-grained proglacial sediments and loamy till interbedded with glaciofluvial and glaciolacustrine facies. Our methodology combines information about sediments in the subsurface with conceptual models of unit relationships to define the positions of bounding surfaces and morphological characteristics of hydrologic facies. Because we are working in a near-surface setting (i.e., depths less than 200 feet), we are able to constrain the conceptual model with bounding surfaces using multiple data sources, such as borehole lithologic information from water well records and rotosonic cores, natural gamma-ray log data, shallow geophysical surveys, and interpreted cross sections.

The 3-D model of the Huntertown Formation is built by constructing georeferenced GIS layers representing the morphology of each major bounding surface. The upper boundary of the model is the ground surface, and the bottom of the model is the top of the overconsolidated glacial till of the Trafalgar Formation (Quaternary). A previously developed two-dimensional geologic map (Fleming, 1994) guides the lateral shape of each internal unit, whereas the morphology on the bottom surface of the model guides placement, thickness, and position of each unit. The geometries of internal units are refined by the conceptual model of depositional relationships (e.g., onlapping, offlapping, interbedded, and so on), especially in areas with sparse data.

The model is being tested and calibrated using well-log and gamma-log interpretations, georeferenced interpreted cross sections, modeled geophysical cross sections, and recent surficial geologic mapping.